摘要:
Video equipment, e.g. for TV producers studios, need to display multiple channels of bit-streams simultaneously. Decoding multiple streams of high bit-rate videos is extremely computational intensive. An improved method for browsing an encoded frequency domain video sequence comprises steps of determining a spatial and/or temporal reduction level for browsing a video sequence, determining according to the determined reduction level frames, fields or parts of blocks or macroblocks to be skipped for decoding, and decoding the remaining data to a video with lower spatial and/or temporal resolution. The reduction may include adaptive partial coefficient decoding, motion vector truncation and/or simplified deblocking filtering. The reduction of spatial resolution is advantageously performed in the frequency domain, so that the inverse quantization process and the iDCT processes require less computing.
摘要:
A method for utilizing a CUDA based GPU to accelerate a complex, sequential task such as video decoding, comprises decoding on a CPU headers and macroblocks of encoded video, performing inverse quantization (on CPU or GPU), transferring the picture data to GPU, where it is stored in a global buffer, and then on the GPU performing inverse waveform transforming of the inverse quantized data, performing motion compensation, buffering the reconstructed picture data in a GPU global buffer, determining if the decoded picture data are used as reference for decoding a further picture, and if so, copying the decoded picture data from the GPU global buffer to a GPU texture buffer. Advantages are that the data communication between CPU and GPU is minimized, the workload of CPU and GPU is balanced and the modules off-loaded to GPU can be efficiently realized since they are data-parallel and compute-intensive.
摘要:
The frame following a scene cut is usually coded as an I picture. In CBR encoding, the encoder will try to keep the bit rate constant, which will often cause serious picture quality degradation at scene changes. In VBR encoding, more bits will be allocated to the first frame of the new scene and the bit rate will increase significantly for a short time. Therefore subsequent frames must be coded in ‘skipped’ mode, which will often cause jerk artifacts. According to the invention, in each frame belonging to a scene change period, areas are determined that have different human attention levels. In the frames (n−1, n−2, n−3) located prior to the first new scene frame, to the areas having a lower attention level less bits are assigned than in the default encoding, and in the frames (n, n+1, n+2) located at and after the scene cut the thus saved bits are additionally assigned to the areas having a higher attention level.
摘要:
Constant-bitrate rate control is very useful for real-time video transmission. However, it is difficult to realize a good CBR rate control in an MPEG4-AVC video codec. According to the invention, an adapted ρ-domain rate control processing for real-time rate control is applied, whereby ρ is the percentage of zero amplitude transform coefficients following quantization. A couple of tables are used for mapping between the quantization parameters and ρ. A detailed analysis at transform level is carried out in order to get the slope of the relationship between the bitrate and ρ. The slope of the linear model is constrained by a look-ahead procedure with respect to the neighbor frames or blocks.
摘要:
Video equipment, e.g. for TV producers studios, need to display multiple channels of bit-streams simultaneously. Decoding multiple streams of high bit-rate videos is extremely computational intensive. An improved method for browsing an encoded frequency domain video sequence comprises steps of determining a spatial and/or temporal reduction level for browsing a video sequence, determining according to the determined reduction level frames, fields or parts of blocks or macroblocks to be skipped for decoding, and decoding the remaining data to a video with lower spatial and/or temporal resolution. The reduction may include adaptive partial coefficient decoding, motion vector truncation and/or simplified deblocking filtering. The reduction of spatial resolution is advantageously performed in the frequency domain, so that the inverse quantization process and the iDCT processes require less computing.
摘要:
The invention is related to spatial diffusion in images. Spatial diffusion helps blurring small discontinuities. Edges become sharper by spatial diffusion as well.A method for generating a spatial diffused image from an input image is described wherein the method comprises applying on the input image an inverted Gaussian spatial bilateral filter with a spatial weight which takes the form of an inverted Gaussian.The inverted-Gaussian-spatial bilateral filter, which uses an inverted-Gaussian function as the kernel of spatial filter, can remove small spots in large smooth areas efficiently.
摘要:
The invention is related to a method for decoding an image frame of an image frame sequence and to a corresponding decoder. The method comprises the steps of reconstructing a representation of the image frame, classifying at least one block boundary as to be either high-freq or low-freq and filtering pixels of the at least one block boundary in dependency on the classification and outputting the filtered representation. Film grain noise and/or random texture are more likely to produce high-freq boundaries. Therefore, filtering boundary pixels in dependency on frequency classification helps preserving film grain noise and random texture which are of relevance for video quality at high bit rates.
摘要:
The frame following a scene cut is usually coded as an I picture. In CBR encoding, the encoder will try to keep the bit rate constant, which will often cause serious picture quality degradation at scene changes. In VBR encoding, more bits will be allocated to the first frame of the new scene and the bit rate will increase significantly for a short time. Therefore subsequent frames must be coded in ‘skipped’ mode, which will often cause jerk artifacts. According to the invention, in each frame belonging to a scene change period, areas are determined that have different human attention levels. In the frames (n−1, n−2, n−3) located prior to the first new scene frame, to the areas having a lower attention level less bits are assigned than in the default encoding, and in the frames (n, n+1, n+2) located at and after the scene cut the thus saved bits are additionally assigned to the areas having a higher attention level.
摘要:
The invention is related to spatial diffusion in images. Spatial diffusion helps blurring small discontinuities. Edges become sharper by spatial diffusion as well.A method for generating a spatial diffused image from an input image is described wherein the method comprises applying on the input image an inverted Gaussian spatial bilateral filter with a spatial weight which takes the form of an inverted Gaussian.The inverted-Gaussian-spatial bilateral filter, which uses an inverted-Gaussian function as the kernel of spatial filter, can remove small spots in large smooth areas efficiently.
摘要:
A method and a device for encapsulating a media entity containing more than one layer into multiple component files, each for one layer, are described along with the corresponding method and device for component file reading. Extensions to the Extractor data structure of SVC/MVC file formats are proposed. The extractor extensions of the invention enable NAL units referencing across different component files. The present invention enables adaptive HTTP streaming of media entities.